Tony Heller's Proxy Trick

Recently I've seen people posting the following graph from Tony Heller, which is intended to show a cooling trend in temperatures following the 1940s and suggest that the "start of the satellite record" begins when temperatures are at their low point. Heller's graph is unsourced except for being called "Briffa's Reconstruction."

This graph has shown up on Twitter, and so far I haven't seen any one identify the source of this data or even what the data represents. However, I was able to determine that up through 1960, the data is identical to Briffa 2001[1], and the data is on an Excel spreadsheet on an FTP site maintained by NOAA for plate 3 of Briffa's paper. The same site also has regional data that includes years following 1960 (plate 2 in Briffa's paper), but Briffa is clear that the post-1960 data does not have a clear temperature signal. Briffa writes, "The period after 1960 was not used to avoid bias in the regression coefficients that could be generated by an anomalous decline in tree density measurements over recent decades that is not forced by temperature [Briffa et al., 1998b]." The divergence problem has been well-documented, especially in Briffa et al 1998; what Briffa is describing here in 2001 was well-understood. Furthermore, Briffa's data is neither global nor annual; it corresponds to the growth season of the Northern Hemisphere land, essentially from April through September. So beginning in 1850, an honest graph of Briffa 2001 should look like this.

The post-1960 data shouldn't be used because the tree ring density measurements in that data are not forced by temperature, so it's not a good proxy for NH growth season temperatures. Of course, this is exactly why Heller used the post-1960 data; he's not interested in showing data with a clear temperature signal. He's interested in showing what will give the impression of cooling following 1940. NOAA's website allows you to download April - September NH land temperatures. If we include both Briffa 2001 through 1960 and NOAA through 2023, this is what the temperature record looks like since 1850.

And of course the most interesting part of Briffa's paper is not the data following 1850 (we have the instrumental record for that) but the data preceding 1850. It turns out that if you plot data from Briffa 2000 and Briffa 2001 along with the instrumental record, you get a hockey stick shape that agrees pretty closely with the MBH99 hockey stick. I show all of these below.

So it would seem that Heller is up to his old tricks again. He's being extremely selective with his use of data, he's deliberately using data that lacks a clear temperature signal as if it does, and he's ignoring the temperature signal recorded by thermometer data. None of the actual data agrees with him, so he has to resort to tricks to create false impression of what's actually going on.

References:

[1] Briffa et al. (2001) J Geophys Res 106, 2929-2941. The data through 1960 is here: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/treering/reconstructions/n_hem_temp/briffa2001jgr3.xls